Plasma triglyceride concentration is an independent although relatively weak risk factor for coronary heart disease (CHD). The relative weakness of plasma triglycerides to predict CHD may be due to the substantial diversity of lipoprotein particles that carry the triglycerides, some being related to atherosclerosis and CHD more than others. We have shown in patients who have had a myocardial infarction that the rather weak association between triglycerides and subsequent coronary events is secondary to a stronger relationship with specific types of VLDL remnants, those in the LDL density range that contain apoCIIl. This application seeks to extend these findings to initial coronary events in patients who do not have CHD. The Primary Specific Aim will evaluate VLDL and LDL particle types as predictors of initial coronary events in men from the Health Professional Follow-up Study (HPFS) and women from the Nurses' Health Study (NHS). A prospective nested case-control design will be used with a total of 1000 CHD cases and 1000 matched controls, with equal numbers of men and women. We will specifically investigate the role of apoCIII containing VLDL and LDL particles in diabetes by over sampling so that 50% of the patients will have type 2 diabetes mellitus. Our previous work shows that LDL apoCIII particles are independent predictors of recurrent CHD in diabetic patients who survived a myocardial infarction. We hypothesize that apoCIII may have a special role in dyslipidemia and CHD in diabetes. Secondary Aims: Besides apoCIII, other small apolipoproteins, apo C1, CII, and All are components of VLDL and LDL and modulate the metabolism of apoB lipoproteins. It is likely that these apolipoproteins have a relationship with human atherosclerosis. We will measure these apolipoproteins in VLDL and LDL and evaluate their relationship to CHD. We will also investigate the associations between these new lipoprotein risk factors and intake of foods and nutrients, physical activity, and other risk factors, including smoking, BMI, age and gender. The results will provide new means to identify nondiabetic and diabetic persons who are at high risk of developing CHD and the environmental determinants, and could form the basis for new lipoprotein targets for lipid management by diet and medicines.